[Boost-testing] Re gression test problem - compiler running 'forever'

Subject: [Boost-testing] Re gression test problem - compiler running 'forever'
From: Richard Webb (richard.webb_at_[hidden])
Date: 2009-05-17 10:12:56

Over the past few days i've been having a problem with some of the Spirit 2
regression tests causing Mingw 4.4.0 (the build from
http://www.tdragon.net/recentgcc/) to get stuck - the compiler sits using
100% cpu without appearing to do anything.

I started a clean run of the Trunk tests earlier, and when i checked them
later on i found 14 instances of g++.exe/cc1plus.exe running, some of which
had been running for over an hour.
This was when running the tests with -j3 and a 40 minute timeout, on Windows
XP.

I don't know what's causing the compilation issue, but it looks like the
test system is running up more copies of the compiler after a test has run
for too long, but isn't terminating the previous process first.

Anyone got any ideas?

Thanks,
Richard Webb

-- 
View this message in context: http://www.nabble.com/Regression-test-problem---compiler-running-%27forever%27-tp23583723p23583723.html
Sent from the Boost - Testing mailing list archive at Nabble.com.